Research On Absolute Positioning Technology Of Tandem Manipulator Based On Binocular Vision

Due to the influence of machining precision,assembly error,transmission error,wear and environmental factors,the robotic arm has low absolute positioning accuracy.With the continuous deepening of the degree of factory automation,higher requirements are imposed on its absolute positioning accuracy.Therefore,it is of great significance to deeply study the absolute positioning of the robotic arm.At present,there are many methods for absolute positioning of a robotic arm,binocular vision has the characteristics of high positioning accuracy and low cost.Therefore,the FANUC robotic arm is used as the research object in this paper,and binocular vision technology is used to realize the high-accuracy absolute positioning of the end-effector of industrial manipulator.The main research work is as follows:Firstly,a new method of elliptical contour extraction is proposed.After the filtering and sharpening of the images collected by the binocular system,the sub-pixel edge extraction algorithm is used to extract the edges of the quality-improved images.And then by limiting the circularity,convexity,eccentricity and the area size to the closed edges of the images,the contour of high reflective markers in the images is extracted effectively.Secondly,the method of camera calibration used in this paper is proposed.Taking full consideration of the first three-order radial distortions and the first two-order tangential distortions of the lens,the Zhang's camera calibration method and the elliptical contour extraction algorithm are combined to complete the calibration of the camera internal parameters and the relative position of the left and right cameras,and using the image residual mean method to verify the validity of the calibration algorithm.Thirdly,according to the specific research of this paper,the epipolar constraint method is used to accomplish the stereo-matched of high reflective markers in the three-dimensional target images.In order to improve the matching efficiency,the RANSAC algorithm is used to effectively eliminate the mis-matching points in the images.The least squares method is used to obtain the 3D world coordinates for the well-matched points,and the accuracy of 3D reconstruction was verified by using a circular array calibration plate that known the distance of the circles centers.By comparing the measured value and the theoretical value,verified the feasibility of the algorithm that used.Finally,a robotic arm measurement system based on binocular tracking was realized.Firstly,the 3D reconstruction of all high reflective markers on the three-dimensional target was accomplished by using the photogrammetry system.And using RANSAC algorithm to match the common markers on the three-dimensional target images captured by the photogrammetry system and the binocular system.And then calibrate the laser and threedimensional target that at the end of the robotic arm to realize the robotic arm measurement system.Then using the system to photograph a scale-plate that known length to verify the absolute positioning accuracy of the robotic arm.By processing and analyzing the three groups of experiments,it shows that the use of binocular vision technology can achieve higher absolute positioning accuracy of the end-effector of industrial manipulator.